You can automate the classifiers in your pyproject.toml file like this:

# [[[cog
# import cog
# import cogeol
#
# for version in reversed(cogeol.scientific()):
#     cycle = version["cycle"]
#     cog.outl(f'  "Programming Language :: Python :: {cycle}",')
# ]]]
"Programming Language :: Python :: 3.11",
# [[[end]]]

Now run the following from the command line:

> cog -c -r pyproject.toml

and you should see the following (notice all versions are present!):

# [[[cog
# import cog
# import cogeol
#
# for version in reversed(cogeol.scientific()):
#     cycle = version["cycle"]
#     cog.outl(f'  "Programming Language :: Python :: {cycle}",')
# ]]]
"Programming Language :: Python :: 3.11",
"Programming Language :: Python :: 3.12",
"Programming Language :: Python :: 3.13",
# [[[end]]] (sum: FeG7grp2Dw)

Let's assume you have the following code snippet in github-workflow.yml:

...
jobs:
  tests-reusable:
    strategy:
      matrix:
        python:
          #
          #           DO NOT EDIT UNTIL end marker
          #
          # [[[cog
          # import cog
          # import cogeol
          #
          # for version in reversed(cogeol.scientific()):
          #     cycle = version['cycle']
          #     cog.outl(f'          - "{cycle}"')
          # ]]]
          - "3.11"
          # [[[end]]] (sum: l3d2zGv79j)

in addition to your code in pyproject.toml using cogeol.

Now, if you run:

> cog -c -r pyproject.toml github-workflow.yml

The following will happen:

• Both files will be updated with appropriate Python versions
• Only one call to End of Life Date will be made (the results are cached on disk)

Next time you run the same command, the results will be read from the cache

For more examples check out this project's:

• pyproject.toml file (see here)
• Tests of the last three versions in GitHub Actions workflow (see here)